Project Summary/Abstract Antibiotic resistance is one of the most serious medical challenges of our time. This crisis puts patients at risk of untreatable bacterial infections and threatens major advances of modern medicine that rely on antibiotics (transplants, chemotherapy, etc). There are at least 2 million antibiotic resistant infections each year in the US, leading to over 23,000 deaths. It is estimated that without significant action, worldwide annual mortality due to these infections will reach 10 million by 2050, surpassing the predicted mortality from cancer. Unfortunately, some bacteria, including specific isolates of carbapenem-resistant Enterobacteriaceae (CRE) and carbapenem-resistant Acinetobacter baumannii (CRAB), are now resistant to all available antibiotics and are essentially untreatable. In such instances, combinations of antibiotics are employed to try to overcome the resistance to individual drugs, but are only sporadically effective. When and why combinations work is unclear, and clinicians therefore lack a sound scientific rationale for choosing antibiotics to include in these regimens. Our research has revealed an unexpected principle, distinct from antibiotic synergy, that can be used to design personalized combinations to kill clinical bacterial isolates including pan-resistant strains. This combination therapy approach is based on heteroresistance, an enigmatic form of antibiotic resistance in which a bacterial isolate harbors a resistant subpopulation that can rapidly replicate in the presence of an antibiotic, while the majority susceptible population is killed. However, we now show that when combined, two antibiotics to which a given strain is heteroresistant, kill the bacteria as each drug inhibits the subpopulation of cells resistant to the other [Band et al, Nature Microbiology, 2019]. Thus, heteroresistance towards multiple antibiotics (?multiple heteroresistance?) can be exploited as a bacterial Achilles' heel and the basis of effective combination regimens. Importantly, this method employs existing FDA-approved antibiotics and can be employed in the clinic immediately. This paradigm-shifting approach to combination therapy has the potential to have a major translational impact, but must first be broadly and thoroughly interrogated. Here, we propose to use a robust set of CRE and CRAB clinical isolates from a Georgia-based surveillance initiative to test for heteroresistance to a wide range of antibiotics. This will allow the selection and in vitro and in vivo testing of combinations targeting multiple heteroresistance. We will further study the relationship between the resistant subpopulations in multiple heteroresistant isolates, as well as performing dynamic flow experiments to determine the pharmacokinetics and pharmacodynamics of effective combinations. This research has the potential to provide clinicians with a rational and predictable method with which to prescribe effective antibiotic combinations to treat bacterial infections, including those currently considered untreatable.